1. Technical Field
The present disclosure relates generally to electing a network master device upon a network failure.
2. Description of the Background Art
Various redundancy schemes are used to prevent network failure. Generally, a network includes at least one master network device and a back-up network device. The back-up network device supports the network when the master network device fails. The network devices interface through various protocols, called “failover protocols,” to detect failure of network connectivity. A heartbeat protocol is one such protocol used for the detection of failed master network device and for activation of back-up network device to take over the functions of the master network device. Conventional methods and systems for redundancy schemes have two paired network devices. For example, two Cisco CSS (Content Service Switch) or two Cisco CSM (Content Switching Module) devices can be paired, to provide redundant network architecture within a data center network. Various multiple deployment modes are available to pair network devices. Active-Active is another mode of pairing devices in which both the devices are simultaneously active in case of failure of a network device.
However, existing methods and systems for redundancy schemes in active-active mode may lead to a dual master condition in which both of the active devices act as master devices. This dual master condition can occur, for example, when one active device is master for internal-facing Local Area Network (LAN) and the second active device is master for external-facing LAN, -based on protocols such as hot standby router protocol (HSRP). The dual-master conditions may also be aggravated by, for example spanning-tree convergence issues in the respective switched LANs, or by momentary connectivity failures. The existence of a dual master (or more than two masters) condition is highly undesirable as it could cause severe network problems. Other network conditions may cause another undesirable situation where no device is acting as the master. These undesirable conditions may be more likely to occur during a failover recovery attempt, or as a result of improper failover recovery.